The ROI of ROI
In: Columns > The $ & Sense of IT
Published on March 13, 2003
So you’ve thought about calculating the ROI of a proposed change to your Web site or for proposing upgrades to servers, software, hardware, or anything else that has to do with your project. When facing this daunting task, you choose to skip it because it's too time consuming. Perhaps you want to measure the ROI of a change you made, and you’re either lost at how to do it or overwhelmed at the time it will take to track and measure it.
It isn’t always easy to put down and quantify on paper what you know in your head and in your heart: “Yes, it will have a positive impact on the bottom line.” What happens if you’ve done the homework in the proposal stage, and you become accountable for it? How are you supposed to measure it? How much time should you invest in tracking every financial aspect of something just to say, “See it was a good recommendation”?
These questions are faced every day. Some people pick up the gauntlet and proceed to propose a ROI and then attempt to track it, while others choose to avoid the whole ROI thing. The reality is, if companies expect their clients (those who hold the purse strings) to continue to invest in Web development, they need to embrace the concept of ROI and learn how to measure it effectively.
So let’s start with a simple equation that every first year accounting student would simply say “Duh!” to. It’s very basic.
- How much is it going to cost (initial investment and on-going costs)?
- How much money is it going to make?
- How much money is it going to save in the first year and every year after that?
From here, figure the ROI. The formula is:
($Earnings + $Savings - $Cost)/$Cost X 100 = ROI %.
Formulas are one thing—take this easy example. You want to buy a computer for working on the company’s Web site. The cost of the new computer with all appropriate bells and whistles is $1,500. In this example, there is nothing wrong with your old computer, it simply isn’t fast enough. So as part of the proposed ROI, we need to determine what is to be done with the old computer. In this example, the company sells it to someone for $500. Hence, the net cost of the new computer is $1,000.
Now you need to determine how much faster the new computer is. Yes, those processor speeds are wonderful guides, but be realistic. If you double the processor speed, you don’t double your productivity. Before you start your ROI calculation, stop and think about how often you’re waiting for the computer to finish processing something before you can proceed.
Where possible keep a paper log beside you and record each instant (both the frequency and duration). This evidence gathering technique will prove very valuable. After a few days, check your log and total the time and occurrences. Divide by the number of days you’ve run your log to come up with the daily nuance factor. Now you know just how much of your time the existing computer is wasting. With this knowledge, you can quickly figure out how much it is costing your employer by multiplying the number of working days in a year and multiplying it by your hourly wage.
For this example, suppose the old computer is costing you on average two minutes 15 times a day for a total of 30 minutes a day. That is approximately 120 hours a year. If you’re earning $15 hour, that amounts to an annual cost of $1,800 a year. Now estimate how much the new computer will save you. Realistically, you’ll never be able to recoup all that time, but let’s assume you can reclaim 50% of it, or $900 in the first year.
In your proposal for the new computer, the boss will want to see a positive ROI, but if you only plug in your numbers for the first year, you’ll get a negative ROI of (900 - (1500-500))/(1500-500)*100 = -10.0%.
Why should the company buy you a new computer? Because the computer has a life expectancy of at least 2 years on your desk, and those savings continue in the second year. So the real ROI in this case needs to cover the second year as well. With this in mind, show your boss a two year ROI projection of ((900+900 - (1500-500))/(1500-500)*100 = 80.0%. Now, what boss in his right mind won’t approve an expenditure that yields a return of 80% in just two years?
Remember that log you were keeping? Well, guess what, some bosses only look at the short term. “I want my payback in the first year.” If this is the case, ask the boss if reducing your frustration from 15 times a day to perhaps seven or eight is worth $100 dollars over the course of a year? Of course it is, and, with that scenario, your ROI is 0 for the first year, which is break even, and that could be enough to get you a new computer.
From this example, reduced frustration has a value. What that value is depends on the individual and is extremely hard, if not impossible, to quantify. You must account for it. As in the above demonstration, while not entirely quantified, it can be the difference of getting something approved or not approved.
While the example of purchasing a new computer may seem a bit trite for taking the time to develop a log and write a proposal quoting a ROI for a purchase of $1,500, the methodology holds true for any investment. Need a new server, monitor and log how often the old one hangs or crashes or simply can’t keep up with demand.. Next, calculate lost revenue due to down time and servicing costs. Don’t forget the impact on customer satisfaction as an intangible cost that the old and inadequate server is generating.
While use of ROI is an effective way to talk to the suits and get what you want, it can also be just as effective in preventing the suits from doing something you don’t want. There are too many companies who make decisions on the fly, not taking into account all the variables. Call it “gut-feel ROI.” Symptoms of this include, but are not limited to, the following phrases:
- I saw this at a trade show. I don’t know how it exactly works, but we need to implement it immediately.
- Can’t we build cheaper than buying it?
When faced with this challenge, buy yourself some time and do your home work. Estimate the implementation time, the impact on the current development schedule, and the subsequent effect on future revenues. If you can demonstrate a quantifiable negative ROI, people will stop and listen. Even if they don’t follow your recommendations, you’ll be comforted when it doesn’t work out that you have an “I told you so” ready to use the next time you’re faced with this situation.
With all this talk about why and how to calculate ROI, what is the ROI of ROI? How much time should one invest in measuring and capturing ROI?
Remember that log you maintained to justify a new computer? How much time did it truly take to develop this indisputable evidence? Would it be worth while to maintain the log once you have the new computer? Of course it would. Why, because the next time you need to justify something, you can demonstrate the actual savings that a previous recommendation has generated. Is it realistic that the log be maintained for entire life of the computer? Of course not. Keeping the log going for a period equal to the original sampling period is highly advised. Compare the actual ROI with the original proposed ROI. The results may surprise you, and when you’ve underestimated that ROI, be sure to let the boss know.
When it comes to larger scale initiatives, the task of knowing what to track and how track it becomes equally larger and complex. When it comes to Web development, there is a lot of evidence waiting in those access logs from the Web servers.
Any average-to-good log analyzing tool is capable of helping to drill down to the details required to deliver an effective ROI. Want to justify costs for redesigning the home page? Find out how many visitors abandon the site after accessing the home page only. Then make your changes and monitor the improvement. Keep tweaking the page and monitoring it until the incremental cost of checking that log file is more then cost of abandonment. For this, you’ll need to quantify what is it worth to the company for each additional customer to proceed beyond the home page.
On an e-commerce site, if improvements translate into more sales then it’s easy to quantify. But what if your site doesn’t sell anything directly? Then you need to dig a little deeper. Do you have a FAQ section or other self-serve information for your customers on the site? If yes, does a more effective Web site mean less calls to your customer support personnel? Of course it does, and this means savings in either reduced labor costs or better customer service.
The reality is that the ROI can be measured for everything. Practicality dictates that you should only measure what you can effectively, to the extent that getting a more detailed measurement is going to cost more than the potential savings. If everyone in the Web world had been doing their ROIs back in the late 90s, perhaps the bubble wouldn’t have occurred, and it wouldn’t have imploded, taking many valuable and potentially successful enterprises with it.
Alan K'necht operates K'nechtology Inc., a search engine optimization and marketing and web development company. He is also a freelance writer, project manager, and accomplished speaker at conferences throughout the world. When he’s not busy working, he can be found chasing his small children or trying to catch some wind while windsurfing or ice/snow sailing.